| Summary: | The Zero-Cast Wx wrist brace is an adjustable orthotic device intended to stabilize distal radius (wrist) fractures (DRF). Designed and developed by the Auckland-based start-up company, Surgisplint Ltd, this patented stabilization system immobilizes the fracture while still retaining a significant degree of joint function - a feature not offered by traditional plaster and fibreglass casts. Additionally, the Zero-Cast Wx brace is a lightweight, foam-lined, plastic injection moulded orthosis that is waterproof and breathable, unlike conventional plaster and fibreglass casts. The current Zero-Cast Wx fitting protocol requires clinician training to account for variations in hand and wrist morphology across patients. Furthermore, localized swelling may also affect the state of the fit after initial application. Clinicians require that patients be incentivized to comply with early stage hand rehabilitation which has been shown to mitigate healing complications and improve functional outcomes of the hand and wrist. This project sought to determine if embedded smart technology in the Zero-Cast Wx brace could be used to assist the fitting procedure. Using structured engineering design methodology, the viability of using embedded smart technology to predict various hand rehabilitation exercises and functional hand tests was also investigated. In formation provided by this smart technology could be used to incentivize patient adherence to hand physiotherapy and also inform the clinician on patient rehabilitation progress. Physiological analysis of the hand and forearm enabled identification of suitable locations for a force sensing array to be placed within the confines of the pre-existing Zero-Cast Wx brace to use Force Myography (FMG). This non-invasive technique detects force patterns produced by changes in limb contour when various underlying muscles are contracted or relaxed. The force patterns correspond to a unique movement or position of the limb relative to a default state, which can be used to train a machine learning algorithm. This method could also be used to classify fitting force in the brace. To demonstrate the feasibility of using embedded sensors in the brace, a machine learning algorithm was trained with FMG data from a single participant to wirelessly classify four hand exercises and a default relaxed position with an accuracy of 92% using a proof-of-concept embedded force sensing system in Zero-Cast Wx. Having proven the viability the of smart system, a pilot study involving 21 participants investigating the use of multi-user FMG training data to classify fitting force, hand mobility exercises, and functional hand tests was undertaken. This study resulted in a classification accuracy of 75% and above for fitting force and hand mobility exercises. Functional testing involving rotating a knob, squeezing a hand exercise ball, and lifting a water bottle resulted in poor classification accuracy. This project has demonstrated the viability of using embedded smart technology in the Zero-Cast Wx brace system to better inform the clinician on brace fitting quality and on patient recovery progress, which may also incentivize patient adherence to hand rehabilitation programs. It was also identified that better classification of functional tests, and possibly other mobility exercises could be achieved through the addition of an inertial measurement unit to the smart brace system.
|
|---|
